This application addresses broad Challenge Area (06) Enabling Technologies and specific Challenge Topic, 06-MH-103: New Technologies for Neuroscience Research. This project will provide a powerful, content-driven approach to the identification of brains having similar geometry and shape, the clustering of neuroanatomically similar cases, and the interactive 3D visualization of the large collections contained in neuroimaging archives. Beginning with the example of the LONI Image Data Archive (IDA), we will design automated data processing meta-workflows that will decompose the thousands of whole brain MRI volumes into constituent 3D neuroanatomical regions. We will characterize the geometric properties of these regional parcellations, store these measurements, and systematically assess pair-wise regional "distances" between brains, and decompose the resulting similarity matrix using multidimensional scaling and related approaches. These processes will be automated to accommodate the continuous growth of the archive and be able to include content obtained from other neuroimaging archives as well. We will graphically represent the derived space of brain similarity via an interactive and freely available 3D browser. Finally, we will develop means for users to upload their own MR anatomical volumes for automated processing via this same process using a large grid computational architecture;decompositions of the uploaded data will be compared against the shape statistics derived from the previously processed archival data;content- based search results will be returned to users via the web in the form of a rank ordered list of brain volumes having similar neuroanatomical characteristics;hyperlinks to additional meta-data and online information, as well as a depiction of the position of their data with respect to other derived brain data using the interactive 3D browser. Meta-data concerning each object in the display will be easily available describing subject demographics, diagnostic group, scanning parameters, etc. This project does not seek to advocate or support the development of any new centralized neuroimaging database but will provide an unprecedented service to the neuroscience community for interacting with existing digital brain archives. The tools developed here will be capable of accommodating that of other neuroimaging repositories as well as user's local archives, thus having utility beyond a single data resource. We expect the outcomes of this project to draw considerable interest and excitement from the neuroimaging community in a similar manner to which BLAST has had for the genomics community. Following a two year development timeline, we anticipate that these informatics-based approaches and tool deliverables will be instrumental for researchers as part of their neuroscientific enterprise, helping to guide research directions, enhance education, and will provide significant new insights concerning large-scale neuroimaging repositories of health and disease. H The outcomes of this novel project for informatics and dynamic visualization to are expected to draw considerable interest and excitement from the neuroimaging community. This project will empower content-driven searches in a similar manner to BLAST has provided for genomics and will provide significant new insights concerning large-scale neuroimaging repositories of health and disease. Over a two-year period of development, during which a number of new American jobs will be created, this project will deliver a robust, content-driven informatics approach to the identification of brains having similar geometry and shape, the clustering of neuroanatomically similar cases, and the interactive 3D visualization of the large collections contained in neuroimaging archives.